Understanding human language

The text of the address by Professor Noam Chomsky at the Special Convocation of the University of Calcutta held on November 22, 2001 to award him the degree of D. Litt Honoris Causa.

I WOULD like to express my gratitude at being welcomed into this distinguished intellectual community, with its vibrant and rich tradition. It is particularly gratifying to receive this honour in India. My professional field, as I am sure you know, was in large part created in India, 2,500 years ago. The first "generative grammar" in something like the modern sense is Panini's grammar of Sanskrit. Nothing was known about these similarities at the origins of the modern versions 50 years ago. It was only after the modern field had taken shape that earlier traditions, long forgotten, began to be explored and reinterpreted in the light of recent insights. Many treasures were discovered, among them Panini's classic - though crucial issues of interpretation remain obscure, and there are surely research topics that could prove highly rewarding.

Meanwhile contemporary inquiries proceeded along their own distinctive course. Characteristically, they view language in a biological setting, adopting what is sometimes called a "biolinguistic approach". From this point of view, the human faculty of language is regarded basically as an organ of the body, mostly the brain, more or less on a par with the visual system or the system of motor organisation.

The language faculty is a "species property" in a dual sense. First, it is close to uniform for the species; second, it is apparently unique to humans in essentials.

The first of these properties is less surprising than it seemed a few years ago. Recent studies indicate that there is remarkably little genetic variation within the human species, so little that it is now commonly assumed that contemporary humans are all descendants of a very small breeding group, perhaps about 100,000 years ago, a conclusion that has broad implications. With regard to language, apart from the margins, variation in the capacity to acquire a rich and highly articulated knowledge of language is so slight as to be virtually undetectable, at least by present means.

The uniqueness property is more surprising. There are no known analogues to human language elsewhere in the animal world. The closest analogies, and these are very weak, are remote: perhaps in some species of insects. Human language does not even find a place in standard taxonomies of animal communication systems; and in fact, there is no strong reason to think of it as primarily a system of communication, contrary to common belief.

Language is like other biological systems, however, in that its basic character is genetically determined. Each person, of course, undergoes a specific course of development, shaped by individual experience, but in highly limited ways. The outcomes are largely a result of shared initial endowment. The human languages, existing or possible, are pretty much cast to the same mould. A rational Martian scientist, studying humans the way we study other animals, could reasonably conclude that there really is only one language, with only minor variations. The variations are very important for our lives; the far deeper uniformities we simply take for granted, without awareness. Similarly, traditional and pedagogical grammars and dictionaries are concerned with the unpredictable and somewhat accidental variation, rightly for their special purposes. The interests of the scientific study of language are virtually complementary: the invariant principles of sound, meaning, and structure that are rooted in our mental nature and that determine the fundamental nature of the languages that each person comes to acquire under normal circumstances.

One basic problem, then, is to discover the invariant principles of the language faculty and the limited options of variation, and then to show that the possible human languages are determined by selecting among the options: one choice yields Tamil (more exactly, a specific variety of Tamil), another yields Swahili, etc. Putting it differently, the task is to show that with a specific choice of options, by adhering to the fixed principles one can literally deduce the infinite array of expressions of the language: their sound, their meaning, the ways in which they can be used to express thoughts, to request information, to tell stories, and numerous others. The task is immensely challenging and difficult. It will doubtless occupy the efforts and energies of many generations of researchers. Nevertheless, there has been quite encouraging progress. In the past 20 years particularly, there has been a flood of discoveries about languages of virtually every known typological variety. Like the well-studied languages, these have been investigated in far greater depth than ever before, revealing many entirely unexpected properties. New questions have arisen that had never been envisioned before. In many cases there have also been plausible answers, sometimes opening new directions for inquiry.

One novel question that has come to the fore in recent years, and that happens to be of particular interest to me, is the question of "optimal design": To what extent is human language an optimal solution to externally-imposed conditions that language must satisfy to be usable at all (for example, accessibility to sensorimotor systems). Equivalently, we may ask to what extent language satisfies the Galilean intuition that "nature is perfect" and it is the task of the scientist to prove it, a guiding intuition for the modern sciences.

There are some answers to such questions for very simple organisms: for example, an explanation of the fact that the shells of viruses are polyhedral. Current work suggests that something similar may be true for human language, a biological system that has emerged in the last moment of evolutionary time, in the most complex organism known, and is surely at the core of our nature and life. Such conclusions, which by now have considerable substance, raise many questions about biological evolution and development generally, and about the human species in particular. These research efforts have also provided new and often surprising solutions to long-standing problems of language, its acquisition and use, and its place in the biological world.

There should be very exciting years ahead in the study of language and other higher mental faculties. There is no better place to pursue these questions than in the land that was the original home of some of the major strands of inquiry that are now being woven into a most intriguing fabric. Speaking personally, I look forward with much eagerness and anticipation to observing, and participating in, these very promising developments.